2005 Annual ReportOur goal is to design new immunotherapies for patients with advanced cancer. Our strategy is based on the use of animal models and human in vitro assays to test hypotheses. We then translate the most promising of these therapies into human clinical trials, which often generate new questions to be tested experimentally. The process is an iterative one that involves close collaboration with basic researchers, biotech scientists and experimental clinicians. Previously, our focus was on the identification of tumor-associated antigens and on the development of vaccines. Despite our extensive clinical and preclinical work with therapeutic cancer vaccines it appears that current cancer vaccines, used alone, are insufficient to reliably induce the objective regression of established cancers in mice or patients. These findings have led us to return to basic efforts to build better vaccines and to use them in combination with other immunotherapeutic modalities. Great progress has been made in the use of vaccination in combination with immunotherapies based on the adoptive transfer of T cells. We have recently made significant advances in identifying the phenotypic and functional qualities that characterize T cells that cause tumor regression. We have also explored the use of ?vC cytokines (especially IL-15 and IL-21) to enhance the antitumor efficacy of vaccine and adoptively transferred cells. Perhaps most significantly, we have very recently found that immunodepletion prior to the adoptive transfer of T cells can greatly enhance therapeutic efficacy. These new immunotherapies have all been extensively modeled in mice. Our ongoing work, although interwoven, has been broken down into three parts in this report:Background - Identifying the antigens - Developing the vaccinesCurrent therapeutic cancer vaccines used alone may be insufficient to reliably induce the objective regression of established cancers in mice or patients. Nevertheless, the identification of new tumor-associated antigens recognized by functional, high-avidity T cells remains important goals. We have identified 5 new human antigenic epitopes: Three HLA class II epitopes derived from proteins known to express class I epitopes include gp100, TRP-1 and NY-ESO-1. These epitopes are presented by class II-expressing melanoma cells and are recognized by human CD4+ T cells. We also identified two other proteins, called OA1 and P polypeptide, that had not been previously known to be targets of an antitumor response in mouse or in man. Although the current cancer vaccines used do not reliably cause tumor regression, recombinant and synthetic immunogens targeting specific tumor-associated antigens can stimulate T cells that specifically recognize tumor targets. Our preclinical efforts have been translated into clinical trials in collaborations for preclinical work and clinical trial design between Drs. Restifo and Rosenberg, together with our clinical team. We have used epitope enhancement, vector improvement (including promoter optimization), cytokines, chemokines and co-stimulatory molecules to enhance therapeutic efficacy of anti-cancer vaccines and we have extensively studied the mechanisms involved in the immunogenicity of new vectors developed in our laboratory, including poxvirus-based vaccines and of replicon-containing "naked" nucleic acid vaccines.Cancer vaccines and adoptive cell transfer-based immunotherapies have been developed and tested by creating new mouse models to study the treatment of large, established cancers Mouse models continue to be useful tools in the development of immunotherapies for patients. We had previously used "model" tumor antigens. Subsequently we identified important murine ortholog genes for human tumor rejection antigens and cloned them into recombinant immunogens.